• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required as is 20 years of participation in forums (not all true). There are daily reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Review and Measurements of Purifi 1ET400A Amplifier

Julf

Major Contributor
Forum Donor
Joined
Mar 1, 2016
Messages
2,363
Likes
3,089
Location
Amsterdam, The Netherlands
Sorry, skipper, just saying stuff like that doesn't "torpedo" anything AFAICT. Like I said, any clipping that was occurring was not readily audible like the examples you trotted out -- yet the tweeters got fried. Interestingly, the 25WPC (Sansui AU-555A) amp's output circuit was capacitance-coupled to the speaker terminals and the the 55WPC (Kenwood KM-8002) replacement was DC-coupled there.
Your anecdotal example is noted. As I wrote, in your case the amp might have been driven into high frequency oscillation.
 

Bruce Morgen

Addicted to Fun and Learning
Forum Donor
Joined
Aug 3, 2019
Messages
632
Likes
801
Your anecdotal example is noted. As I wrote, in your case the amp might have been driven into high frequency oscillation.

It would have to be at a pretty high frequency to be inaudible and pretty powerful to burn out a tweeter. BTW, AFAICT that 25WPC amp was undamaged and I continued to use it as the system's preamp, relegating the power amp section to driving the rear speakers in my old QX 4-channel system. Frankly, I don't think your hypothetical is any more likely than mine -- perhaps someone more astute than I could examine the schematic of the AU-555A to assess the likelihood of our respective hypotheticals. Although we all know that clipping happens to any amp driven hard enough, it would seem that being susceptible to oscillation under the same conditions would be classified as a design defect. Remember, this problem was common enough that the mfr. had to come up with a "Tweeter Saver" accessory, so under your latest scenario any number of amps would have to be prone to that sort of oscillation.
 

Julf

Major Contributor
Forum Donor
Joined
Mar 1, 2016
Messages
2,363
Likes
3,089
Location
Amsterdam, The Netherlands
It would have to be at a pretty high frequency to be inaudible and pretty powerful to burn out a tweeter.
Anything above 20 kHz will do, oscillations can easily be much higher than that, and at the full voltage swing of the amp (so way beyond rated power).
 

peng

Major Contributor
Forum Donor
Joined
May 12, 2019
Messages
2,853
Likes
2,350
Anything above 20 kHz will do, oscillations can easily be much higher than that, and at the full voltage swing of the amp (so way beyond rated power).

Not likely according to the article you linked:
I am sure a 25 W amp can damage some speakers under some conditions but I don't agree with the notion that a bigger amp has less chance to do the same, not as a rule anyway. I assume we would likely agree on this point.

In Bruce Morgen's case, it could also be that his speaker's tweeter(s) voice coil happened to have relatively low inductance so they were more prone to the relatively high order harmonics resulted from clipping as that could result in excessive heat/thermal failure if the high order harmonic contents were high enough.

@Bruce Morgen, did the tweeter(s) fail quickly, or after a long listening session?

8 - Some Very Silly Myths

8.1 - The 'Oscillations' Myth

Since it is possible for an audio amplifier to oscillate at very high ( supersonic ) frequencies, it is widely assumed that this inaudible oscillation will silently damage woofers and instrument speakers in the same way that over powering does. So the story is often trotted out to explain burn voice coils.

FACT: The vast majority of woofers and instrument speakers are immune from damage by inaudible high frequencies. The very high self inductance of the voice coil at frequencies at or above 20 kHz means the current flow is small and no serious heating can happen. Look at the JBL 2226 impedance curve - it's over 100 ohms at 20 kHz [ 1 ].


  1. Damage to tweeters is possible and also to 'twin cone' speakers with copper caps or rings attached to the pole piece.
  2. High frequency oscillation at full power IS damaging to amplifiers. Smoke will appear as resistors and capacitors in output Zobel networks burn up and BJT amps will quickly expire as transistors fail from excess dissipation. Amps using lateral MOSFETs will normally survive such events with only slight damage. The causes of such oscillation lie with operator error and bad cabling practices.
 

DonH56

Master Contributor
Technical Expert
Forum Donor
Joined
Mar 15, 2016
Messages
6,405
Likes
12,208
Location
Monument, CO

AdamG247

Strive to be Understood, rather than just heard!
Moderator
Forum Donor
Joined
Jan 3, 2021
Messages
2,469
Likes
6,119

Julf

Major Contributor
Forum Donor
Joined
Mar 1, 2016
Messages
2,363
Likes
3,089
Location
Amsterdam, The Netherlands

Bruce Morgen

Addicted to Fun and Learning
Forum Donor
Joined
Aug 3, 2019
Messages
632
Likes
801
so hopefully we can move on from an isolated incident that happened 45 years ago... :)

So "isolated" that a major speaker mfr. had to design and offer an accessory literally called the "Tweeter Saver" -- because so many low power 1970s amplifiers went into wildly destructive, inaudibly high frequency oscillation when pushed past their rated output. :facepalm: ;)
 

Julf

Major Contributor
Forum Donor
Joined
Mar 1, 2016
Messages
2,363
Likes
3,089
Location
Amsterdam, The Netherlands
So "isolated" that a major speaker mfr. had to design and offer an accessory literally called the "Tweeter Saver" -- because so many low power 1970s amplifiers went into wildly destructive, inaudibly high frequency oscillation when pushed past their rated output. :facepalm: ;)
I think the article by @DonH56 is a good read, and I have nothing to add to what already has been written by him, me, Phil Allison and Rod Elliott. Let's get back to discussing the Purifi amp.
 

restorer-john

Master Contributor
Joined
Mar 1, 2018
Messages
9,123
Likes
25,772
Location
Gold Coast, Queensland, Australia
I think the article by @DonH56 is a good read, and I have nothing to add to what already has been written by him, me, Phil Allison and Rod Elliott. Let's get back to discussing the Purifi amp.

Don's article is about clipping. It doesn't address the biggest issue with clipping, that it introduces instability (due to the amplifier temporarily running open loop) and oscillation. And it is the HF oscillation that kills the tweeters.

Bruce is right. For those of us who were around selling gear in the early digital days when amplifiers didn't know what had hit them, there were a number of loudspeaker manufacturers who had to come up with tweeter protection really fast or be stuck with an insane number of dubious warranty claims.

And, now, the issue is far from 'solved'. People with low powered amps, be they class 'whatever' and expecting higher levels than the gear can achieve.

Sure, we can hive off a clipping thread, but clipping issues are applicable and can be discussed about any amplifier.
 
Last edited:

restorer-john

Master Contributor
Joined
Mar 1, 2018
Messages
9,123
Likes
25,772
Location
Gold Coast, Queensland, Australia
@Bruce Morgen, did the tweeter(s) fail quickly, or after a long listening session?

I've seen (and had) tweeters expire in an instant from high powered amplifiers that clipped and oscillated.
 

Julf

Major Contributor
Forum Donor
Joined
Mar 1, 2016
Messages
2,363
Likes
3,089
Location
Amsterdam, The Netherlands
Don's article is about clipping. It doesn't address the biggest issue with clipping, that it introduces instability (due to the amplifier temporarily running open loop) and oscillation. And it is the HF oscillation that kills the tweeters.
Indeed. I was really addressing the "(Clipping) low power amps kill speakers, higher powered amps don't" myth. I agree the issue is more likely oscillation. This is relevant to the Purifi discussion in that one of the main things Bruno Putzeys has brought to the class D game is the thorough analysis and understanding of feedback and stability in class D amps.
 

DonH56

Master Contributor
Technical Expert
Forum Donor
Joined
Mar 15, 2016
Messages
6,405
Likes
12,208
Location
Monument, CO
@restorer-john brings up a good point, though again this discussion is way off-topic IMO. I did not consider amplifier (in)stability in my clipping article because I consider it a separate issue. High-frequency oscillation will almost certainly destroy a tweeter, clipped or not, almost instantly. Things like "motor-boating" and other lower-frequency instabilities can blow out other drivers. Heck, a full-power turn-on surge ("pop") can do some damage!

You may not have to overdrive an amplifier to have it oscillate; the wrong load (speaker impedance) can do it. It is also worth noting that clipping is not always as clean as my article indicates; some amplifiers will "fold back" or generate otherwise very ugly asymmetric waveforms (@amirm has shown examples before), and some will break into oscillation when clipped. My article only addresses a well-behaved amplifier that is overdriven resulting in a clipped output. Alas, not all amplifiers are so well behaved.

As @Julf said, part of the work Bruno et. al. did was to develop a thorough understanding of feedback and how it is applied, and misapplied, to make amps like the Purifi models well-behaved.

IME/IME/etc. - Don
 

mk1classic

Member
Joined
Aug 20, 2019
Messages
58
Likes
80
How can we retrace this discussion back to be relevant to the Purifi 1ET400A amplifier?
Do we see a possible problem with HF oscilation with the 1ET400A design? We have seen hf measurements but not under high load?
 

Julf

Major Contributor
Forum Donor
Joined
Mar 1, 2016
Messages
2,363
Likes
3,089
Location
Amsterdam, The Netherlands
How can we retrace this discussion back to be relevant to the Purifi 1ET400A amplifier?
Pretty impossible. :)
Do we see a possible problem with HF oscilation with the 1ET400A design? We have seen hf measurements but not under high load?
No, HF oscillation is definitely not an issue with the 1ET400A (or the earlier Hypex ones for that matter).
 

peng

Major Contributor
Forum Donor
Joined
May 12, 2019
Messages
2,853
Likes
2,350
Pretty impossible. :)

No, HF oscillation is definitely not an issue with the 1ET400A (or the earlier Hypex ones for that matter).

I hope not as I have one each. Such oscillation issue under specific load conditions might have been more common in the old days, I would hope that the newer amps should be virtually immune to that in most cases. I would love to read even one review with measurements such as Stereophile and ASR's that encounter such as issue that caused trouble (including failure) with the amp under test, let alone speakers.
 
Top Bottom